Quality Requirements (QRs) are a key artifact to ensure the quality and success of a software system. Despite its importance, QRs have not reached the same degree of attention as its functional counterparts, especially in the context of trending software development methodologies like Agile Software Development (ASD). Moreover, crucial information that can be obtained from data sources of a project under development (e.g. JIRA, github,…) are not fully exploited, or even neglected, in QR elicitation activities. In this work, we present a data-driven approach to semi-automatically generate and document QRs in the context of ASD. We define an architecture focusing on the process and the artefacts involved. We validate and iterate on such architecture by conducting workshops in four companies of different size and profile. Finally, we present the implementation of such architecture, considering the feedback and outcomes of the conducted workshops.